SAN FRANCISCO, California: OpenAI is set to release GPT-5.6, its most advanced artificial intelligence model, on July 9 after delaying the launch last month at the request of the U.S. government over national security concerns.

The ChatGPT maker said it will launch GPT-5.6 Sol alongside the lower-cost Terra and Luna models, according to a post on X.

The release comes as Washington and Beijing increase scrutiny of advanced AI systems amid concerns they could be misused for military, intelligence or cybersecurity purposes.

The United States has stepped up oversight of cutting-edge AI models over fears that governments in China, Russia and other countries could exploit the technology. Chinese authorities have also been discussing restrictions on overseas access to the country's most advanced AI models.

According to Axios, which first reported the launch plans, the Trump administration approved a broader rollout of GPT-5.6 after additional testing and meetings between OpenAI and government officials.

The White House and the U.S. Department of Commerce did not respond to requests for comment outside regular business hours.

OpenAI had previously limited access to GPT-5.6 to a small group of vetted partners whose identities were shared with U.S. authorities.

When previewing the models in late June, OpenAI highlighted improved agentic capabilities in coding, biology and cybersecurity. It also said GPT-5.6 Sol performed competitively with Anthropic's Mythos Preview on the ExploitBench cybersecurity benchmark.

The rollout follows similar government scrutiny of rival AI developer Anthropic. Last month, Anthropic temporarily disabled its most advanced AI models, Mythos 5 and Fable 5, after a U.S. export control order issued on June 12 over national security concerns. Access was restored last week after the company introduced additional safeguards.

Billionaire Elon Musk said that his company, SpaceXAI, was also making its flagship Grok 4.5 model publicly available.

The developments come as President Donald Trump has introduced a voluntary framework allowing AI developers to provide "covered frontier models" to the U.S. government for up to 30 days before releasing them to trusted partners.

Although export controls on Anthropic's Fable model have been lifted, Mythos, which is designed for cybersecurity professionals, remains available only to certain "trusted" U.S. organizations.

Anthropic has warned it was "probably impossible" to make any AI model fully robust against jailbreaks.